Multi-Point Semantic Representation for Intent Classification

Detecting user intents from utterances is the basis of natural language understanding (NLU) task. To understand the meaning of utterances, some work focuses on fully representing utterances via semantic parsing in which annotation cost is labor-intentsive. While some researchers simply view this as intent classification or frequently asked questions (FAQs) retrieval, they do not leverage the shared utterances among different intents. We propose a simple and novel multi-point semantic representation framework with relatively low annotation cost to leverage the fine-grained factor information, decomposing queries into four factors, i. e. , topic, predicate, object/condition, query type. Besides, we propose a compositional intent bi-attention model under multi-task learning with three kinds of attention mechanisms among queries, labels and factors, which jointly combines coarse-grained intent and fine-grained factor information. Extensive experiments show that our framework and model significantly outperform several state-of-the-art approaches with an improvement of 1. 35%-2. 47% in terms of accuracy.

Authors

• Jinghan Zhang Portland State University
• Yuxiao Ye
• Yue Zhang
• Likun Qiu
• Bin Fu Institute of Computing Technology, Chinese Academy of Sciences University of Chinese Academy of Sciences
• Yang Li
• Zhenglu Yang Nankai University, Tianjin, China
• Jian Sun Xi'an Jiaotong University Pazhou Laboratory (Huangpu), China Peng Cheng Laboratory, China

Keywords

No keywords are indexed for this paper.